The large benefit of the transformer output in a push-pull amplifier is that the current is not limited by the speaker impedance because the primary can have a much lower resistance than the speakers and wiring. That is why the FEDERAL electronic siren could deliver so much power to a ten ohm speaker. A second benefit is that grounding one speaker lead does no damage to the amp. A ground will destry many of the BTL amplifiers.

 

The amplifier with transformers uses antique geranium PNP transistors.

 

Correct, 2N1560.
But it was a workhorse in its day.

 

I have one of those Federal Siren/ amplifiers and the only problem withh those PNP germanium transistors is getting replacements. It was a good device in that era, and with new transistors would still be a good amplifier. NOT HIFI by any stretch but it was never intended to be.

 

Here is a general schematic of the series-AC, parallel-DC push-pull amplifier that I found in my Audio Engineering Handbook. The text says that the peak to peak output voltage across the load is nearly four times the power supply voltage. Peak load current is nearly two times the power supply voltage divided by the load resistance. So assuming a power supply voltage of 12 Volt and load resistance equals 10 Ohm,

output power equals 0.71 *24V *0.71* 2.4A = 29 Watts

or rather approaches being equal to 29 Watts.

From the general schematic given, I certainly wouldn't know how to put together a practical amplifier like this, nor do I completely understand how the voltage and current boosting is derived.

Regards,
Pete

 

That circuit does not go very well with that explanation because the load gets connected to the secondary in a normal circuit, NOT across the primary. Because of the transformer action, when one side of the primary is pulled toward zero volts the other end rises towards twice the V+ value. But a circuit as shown would never be used in a well designed system because both of the output connections are at the V+ potential on the average, and that is just asking for problems.

 

That circuit does not go very well with that explanation because the load gets connected to the secondary in a normal circuit, NOT across the primary. Because of the transformer action, when one side of the primary is pulled toward zero volts the other end rises towards twice the V+ value. But a circuit as shown would never be used in a well designed system because both of the output connections are at the V+ potential on the average, and that is just asking for problems.

The Schematic of the amplifier in my previous post is mostly exactly the way that it is shown in the book. The only difference is that my schematic shows NPN transistors while the transistors in the book are either PNP or NPN (no arrow on the emitter). The section of the book that the amp is discussed in was authored by Daniel von Recklinghausen. While I don't know amplifiers very well at all, I agree that it looks like an odd configuration. That's why I'm bringing it up, I like the odd-ball designs.

Recklinghausen says, "The load may be connected to a secondary winding or between the ends of the primary winding." Then what I said in my previous post occurs when connecting to the primary.

Von Recklinghausen is a graduate of MIT. In what would have been his early days, I suppose, he was Vice President of the H.H. Scott company.

 

I remember, back in the 80's, I produced (to special order) a number of high power siren units to operate on 12V or 24V. They produced about 120watts RMS but were using four of the "new" MOSFET power transistors driving a specially made output transformer. They worked very well although the distortion must have been pretty bad, but that didn't matter because the MOSFETs were driven by a square wave and square waves don't couple through transformers very well. The transformer secondaries fed four special siren re-entrant horns which, I think, had a maximum rating of 50watts each. I suspect that even if you manage to get your requested output power from your DC supply, you may struggle to find speakers that will handle that sort of SQUARE WAVE power - I know I did which was why the output transformer fed FOUR speakers and my units were not capable of ANY speech output.

 

Vs is right about square waves and those folded trumpet speakers. An unexpected failure mode can be voice coil to frame arcing, it seems. I have repaired a couple of them that had slightly burned coils, but afterwards they were 8 ohms and no longer 10 ohm speakers. Still able to produce lots of noise but probably not as reliable. Probably a low resistance low-pass filter could provide a nicer drive waveform.